Llama 3.3 70B Instruct (Meta)

Identity and capabilities

Status: active
Local execution: Available weights · transformers · access approval or licence acceptance required · llama3.3 Hardware requirements and licence restrictions apply.
Context window: 131,072
Maximum output: 16,384
Image input: No
Reasoning: No
Tool calling: Yes
Release date: Dec 6, 2024
First seen: 2026-08-29 19:40:01Z UTC
Last verified: just now

Current offers - prices in USD · per 1M tokens

Prices by provider, service tier, context and offer route
	Amazon Bedrock	Hugging Face Inference Providers ⊕ 4 routes ▾	OpenRouter
Standard		Novita
Cached input	—	—	$0.11
Input	$0.72	$0.135	$0.22
Output	$0.72	$0.4	$0.5

Hugging Face Inference Providers · 4 routes — show all prices

The main table shows Novita: the lowest available starting token price for 2 input tokens per output token. Other input lengths, service tiers and usage ratios may favour another route.

Prices by provider, service tier, context and offer route
	Novita	Ovhcloud	Scaleway	Together AI
Input	$0.135	$0.74	$1.026	$1.04
Output	$0.4	$0.74	$1.026	$1.04

Active promotions

No officially confirmed promotion applies to this model.

Recent price changes

Effective price increased · cause unverified
- Input: $0.1 → $0.22
- Output: $0.32 → $0.5
OpenRoutermeta-llama/llama-3.3-70b-instructdetected 9 Oct 2026, 01:30:09 UTC
Effective price decreased · cause unverified
- Input: $0.22 → $0.1
- Output: $0.5 → $0.32
OpenRoutermeta-llama/llama-3.3-70b-instructdetected 8 Oct 2026, 00:50:20 UTC
Effective price increased · cause unverified
- Input: $0.1 → $0.22
- Output: $0.32 → $0.5
OpenRoutermeta-llama/llama-3.3-70b-instructdetected 5 Oct 2026, 14:07:34 UTC